[432] Curr and Fison were both informed by the same person, D. S. Stewart.

[433] Mathews, Aboriginal Tribes of N.S. Wales and Victoria, in Journal and Proceedings of the Royal Society of N.S. Wales, XXXVIII, pp. 287 f.; Howitt, Nat. Tr., p. 121.

[434] The feminine form of the names given by Mathews is Gurogigurk and Gamatykurk. These are the forms which Howitt reproduces, with a slightly different orthography. The names are also equivalent to those used by the Mount Gambier tribe (Kumite and Kroki).

[435] The native name of this clan is Dyàlup, which Mathews does not translate. This word appears to be identical with Jallup, by which Howitt designates a sub-clan of the same tribe, and which he translates "mussel." That is why we think we can hazard this translation.

[436] This is the translation of Howitt; Mathews renders the word Wartwurt, "heat of the midday sun."

[437] The tables of Mathews and Howitt disagree on many important points. It even seems that clans attributed by Howitt to the Kroki phratry are given to the Gamutch phratry by Mathews, and inversely. This proves the great difficulties that these observations present. But these differences are without interest for our present question.

[438] Mrs. Langloh Parker, The Euahlayi Tribe, pp. 12 ff.

[439] The facts will be found below.

[440] Carr, III, p. 27. Cf. Howitt, Nat. Tr., p. 112. We are merely mentioning the most characteristic facts. For details, one may refer to the memoir already mentioned on Les classifications primitives.

[441] Ibid., pp. 34 ff.

[442] Swanton, The Haida, pp. 13-14, 17, 22.

[443] This is especially clear among the Haida. Swanton says that with them every animal has two aspects. First, it is an ordinary animal to be hunted and eaten; but it is also a supernatural being in the animal's form, upon which men depend. The mythical beings corresponding to cosmic phenomena have the same ambiguity (Swanton, ibid., 16, 14, 25).

[444] See above, p. 142. This is the case among the Gournditch-mara (Howitt, Nat. Tr., p. 124), in the tribes studied by Cameron near the Dead Lake, and among the Wotjobaluk (ibid., pp. 125, 250).

[445] J. Mathews, Two Representative Tribes, p. 139; Thomas, Kinship and Marriage, pp. 53 f.

[446] Among the Osage, for example (see Dorsey, Siouan Sociology, in XVth Rep., pp. 233 ff.).

[447] At Mabuiag, an island in Torrès' Strait (Haddon, Head Hunters, p. 132), the same opposition is found between the two phratries of the Arunta: one includes the men of a water totem, the other those of earth (Strehlow, I, p. 6).

[448] Among the Iroquois there is a sort of tournament between the two phratries (Morgan, Ancient Society, p. 94). Among the Haida, says Swanton, the members of the two phratries of the Eagle and the Crow "are frequently considered as avowed enemies. Husband and wife (who must be of different phratries) do not hesitate to betray each other" (The Haida, p. 62). In Australia this hostility is carried into the myths. The two animals serving the phratries as totems are frequently represented as in a perpetual war against each other (see J. Mathews, Eaglehawk and Crow, a study of Australian Aborigines, pp. 14 ff.). In games, each phratry is the natural rival of the other (Howitt, Nat. Tr., p. 770).

[449] So Thomas has wrongly urged against our theory of the origin of the phratries its inability to explain their opposition (Kinship and Marriage, p. 69). We do not believe that it is necessary to connect this opposition to that of the profane and the sacred (see Hertz, La prééminence de la main droite, in the Revue Philosophique, Dec., 1909, p. 559). The things of one phratry are not profane for the other; both are a part of the same religious system (see below, p. 155).

[450] For example, the clan of the Tea-tree includes the grasses, and consequently herbivorous animals (see Kamilaroi and Kurnai, p. 169). This is undoubtedly the explanation of a particularity of the totemic emblems of North America pointed out by Boas. "Among the Tlinkit," he says, "and all the other tribes of the coast, the emblem of a group includes the animals serving as food to the one whose name the group bears" (Fifth Rep. of the Committee, etc., British Association for the Advancement of Science, p. 25).

[451] Thus, among the Arunta, frogs are connected with the totem of the gum-tree, because they are frequently found in the cavities of this tree; water is related to the water-hen; with the kangaroo is associated a sort of parrot frequently seen flying about this animal (Spencer and Gillen, Nat. Tr., pp. 146-147, 448).

[452] One of the signs of this primitive lack of distinction is that territorial bases are sometimes assigned to the classes just as to the social divisions with which they were at first confounded. Thus, among the Wotjobaluk in Australia and the Zuñi in America, things are ideally distributed among the different regions of space, just as the clans are. Now this regional distribution of things and that of the clans coincide (see De quelques formes primitives de classification, pp. 34 ff.). Classifications keep something of this special character even among relatively advanced peoples, as for example, in China (ibid., pp. 55 ff.).

[453] Bridgmann, in Brough Smyth, The Aborigines of Victoria, I, p. 91.

[454] Fison and Howitt, Kamilaroi and Kurnai, p. 168; Howitt, Further Notes on the Australian Class Systems, J.A.I., XVIII, p. 60.

[455] Curr, III, p. 461. This is about the Mount Gambier tribe.

[456] Howitt, On some Australian Beliefs, J.A.I., XIII, p. 191, n. 1.

[457] Howitt, Notes on Australian Message Sticks, J.A.I., XVIII, p. 326; Further Notes, J.A.I., XVIII, p. 61, n. 3.

[458] Curr, III, p. 28.

[459] Mathews, Ethnological Notes on the Aboriginal Tribes of N.S. Wales and Victoria, in Journ. and Proceed. of the Royal Soc. of N.S. Wales, XXXVIII, p. 294.

[460] Cf. Curr, III, p. 461; and Howitt, Nat. Tr., p. 146. The expressions Tooman and Wingo are applied to the one and the other.

[461] Howitt, Nat. Tr., p. 123.

[462] Spencer and Gillen, Nat. Tr., pp. 447 ff.; cf. Strehlow, III, pp. xii ff.

[463] Fison and Howitt, Kamilaroi and Kurnai, p. 169.

[464] Curr, III, p. 462.

[465] Mrs. Parker, The Euahlayi Tribe, p. 20.

[466] Spencer and Gillen, Nor. Tr., p. 151; Nat. Tr., p. 447; Strehlow, III, p. xii.

[467] Spencer and Gillen, Nat. Tr., p. 449.

[468] However, there are certain tribes in Queensland where the things thus attributed to a social group are not forbidden for the members of the group: this is notably the case with the Wakelbura. It is to be remembered that in this society, it is the matrimonial classes that serve as the framework of the classification (see above, p. 144). Not only are the men of one class allowed to eat the animals attributed to this class, but they may eat no others. All other food is forbidden them (Howitt, Nat. Tr., p. 113; Curr, III, p. 27).

But we must not conclude from this that these animals are considered profane. In fact, it should be noticed that the individual not only has the privilege of eating them, but that he is compelled to do so, for he cannot nourish himself otherwise. Now the imperative nature of this rule is a sure sign that we are in the presence of things having a religious nature, only this has given rise to a positive obligation rather than the negative one known as an interdiction. Perhaps it is not quite impossible to see how this deviation came about. We have seen above (p. 140) that every individual is thought to have a sort of property-right over his totem and consequently over the things dependent upon it. Perhaps, under the influence of special circumstances, this aspect of the totemic relation was developed, and they naturally came to believe that only the members of the clan had the right of disposing of their totem and all that is connected with it, and that others, on the contrary, did not have the right of touching it. Under these circumstances, a tribe could nourish itself only on the food attributed to it.

[469] Mrs. Parker uses the expression "multiplex totems."

[470] As examples, see the Euahlayi tribe in Mrs. Parker's book (pp. 15 ff.) and the Wotjobaluk (Howitt, Nat. Tr., pp. 121 ff.; cf. the above-mentioned article of Mathews).

[471] See the examples in Howitt, Nat. Tr., p. 122.

[472] See our De quelques formes primitives de classification, p. 28, n. 2.

[473] Strehlow, II, pp. 61-72.

[474] Nat. Tr., p. 112.

[475] See especially Nat. Tr., p. 447, and Nor. Tr., p. 151.

[476] Strehlow, III, pp. xiii-xviii. It sometimes happens that the same secondary totems are attached to two or three principal totems at the same time. This is undoubtedly because Strehlow has not been able to establish with certainty which is the principal totem.

Two interesting facts which appear from this table confirm certain propositions which we had already formulated. First, the principal totems are nearly all animals, with but rare exceptions. Also, stars are always only secondary or associated totems. This is another proof that these latter were only slowly advanced to the rank of totems and that at first the principal totems were preferably chosen from the animal kingdom.

[477] According to the myth, the associate totems served as food to the men of the principal totem in the fabulous times, or, when these are trees, they gave their shade (Strehlow, III, p. xii; Spencer and Gillen, Nat. Tr., p. 403). The fact that the associate totems are believed to have been eaten does not imply that they are considered profane; for in the mythical period, the principal totem itself was consumed by the ancestors, the founders of the clan, according to the belief.

[478] Thus in the Wild Cat clan, the designs carved on the churinga represent the Hakea tree, which is a distinct totem to-day (Spencer and Gillen, Nat. Tr., pp. 147 f.). Strehlow (III, p. xii, n. 4) says that this is frequent.

[479] Spencer and Gillen, Nor. Tr., p. 182; Nat. Tr., pp. 151 and 297.

[480] Nat. Tr., pp. 151 and 158.

[481] Ibid., pp. 448 and 449.

[482] Thus Spencer and Gillen speak of a pigeon called Inturrita, sometimes as a principal totem (Nat. Tr., p. 410), sometimes as an associate totem (ibid., p. 448).

[483] Howitt, Further Notes, pp. 63-64.

[484] Thus it comes about that the clan has frequently been confounded with the tribe. This confusion, which frequently introduces trouble into the writings of ethnologists, has been made especially by Curr (I, pp. 61 ff.).

[485] This is the case especially among the Warramunga (Nor. Tr., p. 298).

[486] See, for example, Spencer and Gillen, Nat. Tr., p. 380 and passim.

[487] One might even ask if tribal totems do not exist sometimes. Thus, among the Arunta, there is an animal, the wild cat, which serves as totem to a particular clan, but which is forbidden for the whole tribe; even the people of other clans can eat it only very moderately (Nat. Tr., p. 168). But we believe that it would be an abuse to speak of a tribal totem in this case, for it does not follow from the fact that the free consumption of an animal is forbidden that this is a totem. Other causes can also give rise to an interdiction. The religious unity of the tribe is undoubtedly real, but this is affirmed with the aid of other symbols. We shall show what these are below (Bk. II, ch. ix).

[488] The totems belong to the tribe in the sense that this is interested as a body in the cult which each clan owes to its totem.

[489] Frazer has made a very complete collection of the texts relative to individual totemism in North America (Totemism and Exogamy, III, pp. 370-456).

[490] For example, among the Hurons, the Iroquois, the Algonquins (Charlevoix, Histoire de la Nouvelle France, VI, pp. 67-70; Sagard, Le grand voyage au pays des Hurons, p. 160), or among the Thompson Indians (Teit, The Thompson Indians of British Columbia, p. 355).

[491] This is the case of the Yuin (Howitt, Nat. Tr., p. 133), the Kurnai (ibid., p. 135), several tribes of Queensland (Roth, Superstition, Magic and Medicine, North Queensland Ethnography, Bulletin No. 5, p. 19; Haddon, Head-Hunters, p. 193); among the Delaware (Heckewelder, An Account of the History ... of the Indian Nations, p. 238), among the Thompson Indians (Teit, op. cit., p. 355), and among the Salish Statlumh (Hill Tout, Rep. of the Ethnol. of the Statlumh, J.A.I., XXXV. pp. 147 ff.).

[492] Hill Tout, loc. cit., p. 154.

[493] Catlin, Manners and Customs, etc., London, 1876, I, p. 36.

[494] Lettres édifiantes et curieuses, new edition, VI, pp. 172 ff.

[495] Charlevoix, op. cit., VI, p. 69.

[496] Dorsey, Siouan Cults, XIth Rep., p. 443.

[497] Boas, Kwakiutl, p. 323.

[498] Hill Tout, loc. cit., p. 154.

[499] Boas, Kwakiutl, p. 323.

[500] Miss Fletcher, The Import of the Totem, a Study from the Omaha Tribe (Smithsonian Rep. for 1897, p. 583).—Similar facts will be found in Teit, op. cit., pp. 354, 356; Peter Jones, History of the Ojibway Indians, p. 87.

[501] This is the case, for example, with the dog among the Salish Statlumh, owing to the condition of servitude in which it lives (Hill Tout, loc. cit., p. 153).

[502] Langloh Parker, Euahlayi, p. 21.

[503] "The spirit of a man," says Mrs. Parker (ibid.), "is in his Yuanbeai (his individual totem), and his Yuanbeai is in him."

[504] Langloh Parker, Euahlayi, p. 20. It is the same among certain Salish (Hill Tout, Ethn. Rep. on the Stseelis and Skaulits Tribes, J.A.I., XXXIV, p. 324). The fact is quite general among the Indians of Central America (Brinton, Nagualism, a Study in Native American Folklore and History, in Proceed. of the Am. Philos. Soc., XXXIII, p. 32).

[505] Parker, ibid.; Howitt, Nat. Tr., p. 147; Dorsey, Siouan Cults, XIth Rep., p. 443. Frazer has made a collection of the American cases and established the generality of the interdiction (Totemism and Exogamy, III, p. 450). It is true that in America, as we have seen, the individual must kill the animal whose skin serves to make what ethnologists call his medicine-sack. But this usage has been observed in five tribes only; it is probably a late and altered form of the institution.

[506] Howitt, Nat. Tr., pp. 135, 147, 387; Australian Medicine Men, J.A.I., XVI, p. 34; Teit, The Shuswap, p. 607.

[507] Meyer, Manners and Customs of the Aborigines of the Encounter Bay Tribe, in Woods, p. 197.

[508] Boas, VIth Rep. on the North-West Tribes of Canada, p. 93; Teit, The Thompson Indians, p. 336; Boas, Kwakiutl, p. 394.

[509] Facts will be found in Hill Tout, Rep. of the Ethnol. of the Statlumh, J.A.I., XXXV, pp. 144, 145. Cf. Langloh Parker, op. cit., p. 29.

[510] According to information given by Howitt in a personal letter to Frazer (Totemism and Exogamy, I, p. 495, and n. 2).

[511] Hill Tout, Ethnol. Rep. on the Stseelis and Skaulits Tribes, J.A.I., XXXIV, p. 324.

[512] Howitt, Australian Medicine Men, J.A.I., XVI, p. 34; Lafitau, Mœurs des Sauvages Amériquains, I, p. 370; Charlevoix, Histoire de la Nouvelle France, VI, p. 68. It is the same with the atai and tamaniu in Mota (Codrington, The Melanesians, pp. 250 f.).

[513] Thus the line of demarcation between the animal protectors and fetishes, which Frazer has attempted to establish, does not exist. According to him, fetishism commences when the protector is an individual object and not a class (Totemism, p. 56); but it frequently happens in Australia that a determined animal takes this part (see Howitt, Australian Medicine Men, J.A.I., XVI, p. 34). The truth is that the ideas of fetish and fetishism do not correspond to any definite thing.

[514] Brinton, Nagualism, in Proceed. Amer. Philos. Soc., XXXIII, p. 32.

[515] Charlevoix, VI, p. 67.

[516] Hill Tout, Rep. on the Ethnol. of the Statlumh of British Columbia, J.A.I., XXXV, p. 142.

[517] Hill Tout, Ethnol. Rep. on the Stseelis and Skaulits Tribes, J.A.I., XXXIV, pp. 311 ff.

[518] Howitt, Nat. Tr., p. 133.

[519] Langloh Parker, op. cit., p. 20.

[520] J. W. Powell, An American View of Totemism, in Man, 1902, No. 84; Tylor, ibid., No. 1; Andrew Lang has expressed analogous ideas in Social Origins, pp. 133-135. Also Frazer himself, turning from his former opinion, now thinks that until we are better acquainted with the relations existing between collective totems and "guardian spirits," it would be better to designate them by different names (Totemism and Exogamy, III, p. 456).

[521] This is the case in Australia among the Yuin (Howitt, Nat. Tr., p. 81), and the Narrinyeri (Meyer, Manners and Customs of the Aborigines of the Encounter Bay Tribe, in Woods, pp. 197 ff.).

[522] "The totem resembles the patron of the individual no more than an escutcheon resembles the image of a saint," says Tylor (op. cit., p. 2). Likewise, if Frazer has taken up the theory of Tylor, it is because he refuses all religious character to the totem of the clan (Totemism and Exogamy, III, p. 452).

[523] See below, chapter ix of this book.

[524] Yet according to one passage in Mathews, the individual totem is hereditary among the Wotjobaluk. "Each individual," he says, "claims some animal, plant or inanimate object as his special and personal totem, which he inherits from his mother" (Journ. and Proc. of the Roy. Soc. of N.S. Wales, XXXVIII, p. 291). But it is evident that if all the children in the same family had the personal totem of their mother, neither they nor she would really have personal totems at all. Mathews probably means to say that each individual chooses his individual totem from the list of things attributed to the clan of his mother. In fact, we shall see that each clan has its individual totems which are its exclusive property; the members of the other clans cannot make use of them. In this sense, birth determines the personal totem to a certain extent, but to a certain extent only.

[525] Heckewelder, An Account of the History, Manners and Customs of the Indian Nations who once inhabited Pennsylvania, in Transactions of the Historical and Literary Committee of the American Philosophical Society, I, p. 238.

[526] See Dorsey, Siouan Cults, XIth Rep., p. 507; Catlin, op. cit., I, p. 37; Miss Fletcher, The Import of the Totem, in Smithsonian Rep. for 1897, p. 580; Teit, The Thompson Indians, pp. 317-320; Hill Tout, J.A.I., XXXV, p. 144.

[527] But some examples are found. The Kurnai magicians see their personal totems revealed to them in dreams (Howitt, Nat. Tr., p. 387; On Australian Medicine Men, in J.A.I., XVI, p. 34). The men of Cape Bedford believe that when an old man dreams of something during the night, this thing is the personal totem of the first person he meets the next day (W. E. Roth, Superstition, Magic and Medicine, p. 19). But it is probable that only supplementary and accessory totems are acquired in this way; for in this same tribe another process is used at the moment of initiation, as we said in the text.

[528] In certain tribes of which Roth speaks (ibid.); also in certain tribes near to Maryborough (Howitt, Nat. Tr., p. 147).

[529] Among the Wiradjuri (Howitt, Nat. Tr., p. 406; On Australian Medicine Men, in J.A.I., XVI, p. 50).

[530] Roth, loc. cit.

[531] Haddon, Head Hunters, pp. 193 ff.

[532] Among the Wiradjuri (same references as above, n. 4).

[533] In general, it seems as though these transmissions from father to son never take place except when the father is a shaman or a magician. This is also the case among the Thompson Indians (Teit, The Thompson Indians, p. 320) and the Wiradjuri, of whom we just spoke.

[534] Hill Tout (J.A.I., XXXV, pp. 146 f.). The essential rite is the blowing upon the skin: if this were not done correctly, the transmission would not take place. As we shall presently see, the breath is the soul. When both breathe upon the skin of the animal, the magician and the recipient each exhale a part of their souls, which are thus fused, while partaking at the same time of the nature of the animal, who also takes part in the ceremony in the form of its symbol.

[535] N. W. Thomas, Further Remarks on Mr. Hill Tout's Views on Totemism, in Man, 1904, p. 85.

[536] Langloh Parker, op. cit., pp. 20, 29.

[537] Hill Tout, in J.A.I., XXXV, pp. 143 and 146; ibid., XXXIV, p. 324.

[538] Parker, op. cit., p. 30; Teit, The Thompson Indians, p. 320; Hill Tout, in J.A.I., XXXV, p. 144.

[539] Charlevoix, VI, p. 69.

[540] Hill Tout, ibid., p. 145.

[541] Thus at the birth of a child, a tree is planted which is cared for piously; for it is believed that its fate and the child's are united. Frazer, in his Golden Bough, gives a number of customs and beliefs translating this same idea in different ways. (Cf. Hartland, Legend of Perseus, II, pp. 1-55.)

[542] Howitt, Nat. Tr., pp. 148 ff.; Fison and Howitt, Kamilaroi and Kurnai, pp. 194, 201 ff.; Dawson, Australian Aborigines, p. 52. Petrie also mentions it in Queensland (Tom Petrie's Reminiscences of Early Queensland, pp. 62 and 118).

[543] Journ. and Proc. of the Roy. Soc. of N.S. Wales, XXXVIII, p. 339. Must we see a trace of sexual totemism in the following custom of the Warramunga? When a dead person is buried, a bone of the arm is kept. If it is a woman, the feathers of an emu are added to the bark in which it is wrapped up; if it is a man, the feathers of an owl (Nor. Tr., p. 169).

[544] Some cases are cited where each sexual group has two sexual totems; thus the Wurunjerri unite the sexual totems of the Kurnai (the emu-wren and the linnet) to those of the Wotjobaluk (the bat and the nightjar owl). See Howitt, Nat. Tr., p. 150.

[545] Totemism, p. 51.

[546] Kamilaroi and Kurnai, p. 215.

[547] Threlkeld, quoted by Mathews, loc. cit., p. 339.

[548] Howitt, Nat. Tr., pp. 148, 151.

[549] Kamilaroi and Kurnai, pp. 200-203; Howitt, Nat. Tr., p. 149; Petrie, op. cit., p. 62. Among the Kurnai, these bloody battles frequently terminate in marriages of which they are, as it were, a sort of ritual precursor. Sometimes they are merely plays (Petrie, loc. cit.).

[550] On this point, see our study on La Prohibition de l'inceste et ses origines, in the Année Sociologique, I, pp. 44 ff.

[551] However, as we shall presently see (ch. ix), there is a connection between the sexual totems and the great gods.

[552] Primitive Culture, I, p. 402; II, p. 237; Remarks on Totemism, with especial reference to some modern theories concerning it, in J.A.I., XXVIII, and I, New Series, p. 138.

[553] Het Animisme bij den Volken van den indischen Archipel, pp. 69-75.

[554] Tylor, Primitive Culture, II, p. 6.

[555] Tylor, ibid., II, pp. 6-18.

[556] G. McCall Theal, Records of South-Eastern Africa, VII. We are acquainted with this work only through an article by Frazer, South African Totemism, published in Man, 1901, No. III.

[557] Codrington, The Melanesians, pp. 32 f., and a personal letter by the same author cited by Tylor in J.A.I., XXVIII, p. 147.

[558] This is practically the solution adopted by Wundt (Mythus und Religion, II, p. 269).

[559] It is true that according to Tylor's theory, a clan is only an enlarged family; therefore whatever may be said of one of these groups is, in his theory, applicable to the other (J.A.I., XXVIII, p. 157). But this conception is exceedingly contestable; only the clan presupposes a totem, which has its whole meaning only in and through the clan.

[560] For this same conception, see A. Lang, Social Origins, p. 150.

[561] See above, p. 63.

[562] Primitive Culture, II, p. 17.

[563] Wundt, who has revived the theory of Tylor in its essential lines, has tried to explain this mysterious relationship of the man and the animal in a different way: it was the sight of the corpse in decomposition which suggested the idea. When they saw worms coming out of the body, they thought that the soul was incarnate in them and escaped with them. Worms, and by extension, reptiles (snakes, lizards, etc.), were therefore the first animals to serve as receptacles for the souls of the dead, and consequently they were also the first to be venerated and to play the rôle of totems. It was only subsequently that other animals and plants and even inanimate objects were elevated to the same dignity. But this hypothesis does not have even the shadow of a proof. Wundt affirms (Mythus und Religion, II, p. 296) that reptiles are much more common totems than other animals; from this, he concludes that they are the most primitive. But we cannot see what justifies this assertion, in the support of which the author cites no facts. The lists of totems gathered either in Australia or in America do not show that any special species of animal has played a preponderating rôle. Totems vary from one region to another with the flora and fauna. Moreover, if the circle of possible totems was so closely limited at first, we cannot see how totemism was able to satisfy the fundamental principle which says that the two clans or sub-clans of a tribe must have two different totems.

[564] "Sometimes men adore certain animals," says Tylor, "because they regard them as the reincarnation of the divine souls of the ancestors; this belief is a sort of bridge between the cult rendered to shades and that rendered to animals" (Primitive Culture, II, p. 805, cf. 309, in fine). Likewise, Wundt presents totemism as a section of animalism (II, p. 234).

[565] See above, p. 139.

[566] Introduction to the History of Religions, pp. 97 ff.

[567] See above, p. 28.

[568] Jevons recognizes this himself, saying, "It is to be presumed that in the choice of an ally he would prefer ... the kind or species which possessed the greatest power" (p. 101).

[569] 2nd Edition, III, pp. 416 ff.; see especially p. 419, n. 5. In more recent articles, to be analysed below, Frazer exposes a different theory, but one which does not, in his opinion, completely exclude the one in the Golden Bough.

[570] The Origin of the Totemism of the Aborigines of British Columbia, in Proc. and Transact. of the Roy. Soc. of Canada, 2nd series, VII, § 2, pp. 3 ff. Also, Report on the Ethnology of the Statlumh, J.A.I., XXXV, p. 141. Hill Tout has replies to various objections made to his theory in Vol. IX of the Transact. of the Roy. Soc. of Canada, pp. 61-99.

[571] Alice C. Fletcher, The Import of the Totem, in Smithsonian Report for 1897, pp. 577-586.

[572] The Kwakiutl Indians, pp. 323 ff., 336-338, 393.

[573] The Development of the Clan System, in Amer. Anthrop., N.S. VI, 1904, pp. 477-486.

[574] J.A.I., XXXV, p. 142.

[575] Ibid., p. 150. Cf. Vth Rep. on the ... N.W. Tribes of Canada, B.A.A.S., p. 24. A myth of this sort has been quoted above.

[576] J.A.I., XXXV, p. 147.

[577] Proc. and Transact., etc., VII, § 2, p. 12.

[578] See The Golden Bough,² III, pp. 351 ff. Wilken had already pointed out similar facts in De Simsonsage, in De Gids, 1890; De Betrekking tusschen Menschen-Dieren en Plantenleven, in Indische Gids, 1884, 1888; Ueber das Haaropfer, in Revue Coloniale Internationale, 1886-1887.

[579] For example, Eylmann in Die Eingeborenen der Kolonie Südaustralien, p. 199.

[580] Mrs. Parker says in connection with the Euahlayi, that if the Yunbeai does "confer exceptional force, it also exposes one to exceptional dangers, for all that hurts the animal wounds the man" (Euahlayi, p. 29).

[581] In a later work (The Origin of Totemism, in The Fortnightly Review, May, 1899, pp. 844-845), Frazer raises this objection himself. "If," he says, "I deposit my soul in a hare, and my brother John (a member of another clan) shoots that hare, roasts and swallows it, what becomes of my soul? To meet this obvious danger it is necessary that John should know the state of my soul, and that, knowing it, he should, whenever he shoots a hare, take steps to extract and restore to me my soul before he cooks and dines upon the animal." Now Frazer believes that he has found this practice in use in Central Australia. Every year, in the course of a ceremony which we shall describe presently, when the animals of the new generation arrive at maturity, the first game to be killed is presented to men of that totem, who eat a little of it; and it is only after this that the men of the other clans may eat it freely. This, says Frazer, is a way of returning to the former the souls they may have confided to these animals. But, aside from the fact that this interpretation of the fact is wholly arbitrary, it is hard not to find this way of escaping the danger rather peculiar. This ceremony is annual; long days may have elapsed since the animal was killed. During all this time, what has become of the soul which it sheltered and the individual whose life depended on this soul? But it is superfluous to insist upon all the inconceivable things in this explanation.

[582] Parker, op. cit., p. 20; Howitt, Australian Medicine Men, in J.A.I., XVI, pp. 34, 49 f.; Hill Tout, J.A.I., XXXV, p. 146.

[583] According to Hill Tout himself, "The gift or transmission (of a personal totem) can only be made or effected by certain persons, such as shamans, or those who possess great mystery power" (J.A.I., p. 146). Cf. Langloh Parker, op. cit., pp. 29-30.

[584] Cf. Hartland, Totemism and some recent Discoveries, in Folk-Lore, XI, pp. 59 ff.

[585] Except perhaps the Kurnai; but even in this tribe, there are sexual totems in addition to the personal ones.

[586] Among the Wotjobaluk, the Buandik, the Wiradjuri, the Yuin and the tribes around Maryborough (Queensland). See Howitt, Nat. Tr., pp. 114-147; Mathews, J. of the R. Soc. of N.S. Wales, XXXVIII, p. 291. Cf. Thomas, Further Notes on Mr. Hill Tout's Views on Totemism, in Man, 1904, p. 85.

[587] This is the case with the Euahlayi and the facts of personal totemism cited by Howitt, Australian Medicine Men, in J.A.I., XVI, pp. 34, 35, 49-50.

[588] Miss Fletcher, A Study of the Omaha Tribe, in Smithsonian Report for 1897, p. 586; Boas, The Kwakiutl, p. 322. Likewise, Vth Rep. of the Committee ... of the N.W. Tribes of the Dominion of Canada, B.A.A.S., p. 25; Hill Tout, J.A.I., XXXV, p. 148.

[589] The proper names of the gentes, says Boas in regard to the Tlinkit, are derived from their respective totems, each gens having its special names. The connection between the name and the (collective) totem is not very apparent sometimes, but it always exists (Vth Rep. of the Committee, etc., p. 25). The fact that individual forenames are the property of the clan, and characterize it as surely as the totem, is also found among the Iroquois (Morgan, Ancient Society, p. 78), the Wyandot (Powell, Wyandot Government, in Ist Rep., p. 59), the Shawnee, Sauk and Fox (Morgan, Ancient Society, pp. 72, 76-77) and the Omaha (Dorsey, Omaha Sociology, in IIIrd Rep., pp. 227 ff.). Now the relation between forenames and personal totems is already known (see above, p. 157).

[590] "For example," says Mathews, "if you ask a Wartwurt man what totem he is, he will first tell his personal totem, and will probably then enumerate those of his clan" (Jour. of the Roy. Soc. of N.S. Wales, XXXVIII, p. 291).

[591] The Beginnings of Religion and Totemism among the Australian Aborigines, in Fortnightly Review, July, 1905, pp. 162 ff., and Sept., p. 452. Cf. the same author, The Origin of Totemism, ibid., April, 1899, p. 648, and May, p. 835. These latter articles, being slightly older, differ from the former on one point, but the foundation of the theory is not essentially different. Both are reproduced in Totemism and Exogamy, I, pp. 89-172. In the same sense, see Spencer and Gillen, Some Remarks on Totemism as applied to Australian Tribes, in J.A.I., 1899, pp. 275-280, and the remarks of Frazer on the same subject, ibid., pp. 281-286.

[592] "Perhaps we may ... say that it is but one remove from the original pattern, the absolutely original form of totemism" (Fortnightly Review, Sept., 1905. p. 455).

[593] On this point, the testimony of Strehlow (II, p. 52) confirms that of Spencer and Gillen. For a contrary opinion, see A. Lang, The Secret of the Totem, p. 190.

[594] A very similar idea had already been expressed by Haddon in his Address to the Anthropological Section (B.A.A.S., 1902, pp. 8 ff.). He supposes that at first, each local group had some food which was especially its own. The plant or animal thus serving as the principal item of food became the totem of the group.

All these explanations naturally imply that the prohibitions against eating the totemic animal were not primitive, but were even preceded by a contrary prescription.

[595] Fortnightly Review, Sept., 1905, p. 458.

[596] Fortn. Rev., May, 1899, p. 835, and July, 1905, pp. 162 ff.

[597] Though considering totemism only a system of magic, Frazer recognizes that the first germs of a real religion are sometimes found in it (Fortn. Rev., July, 1905, p. 163). On the way in which he thinks religion developed out of magic, see The Golden Bough,² I, pp. 75-78.

[598] Sur le totemisme, in Année Soc., V, pp. 82-121. Cf., on this same question, Hartland, Presidential Address, in Folk-Lore, XI, p. 75; A. Lang, A Theory of Arunta Totemism, in Man, 1904, No. 44; Conceptional Totemism and Exogamy, ibid., 1907, No. 55; The Secret of the Totem, ch. iv; N. W. Thomas, Arunta Totemism, in Man, 1904, No. 68; P. W. Schmidt, Die Stellung der Aranda unter der Australischen Stämmen, in Zeitschrift für Ethnologie, 1908, pp. 866 ff.

[599] Die Aranda, II, pp. 57-58.

[600] Schulze, loc cit., pp. 238-239.

[601] In the conclusion of Totemism and Exogamy (IV, pp. 58-59), Frazer says, it must be admitted, that there is a totemism still more ancient than that of the Arunta: it is the one observed by Rivers in the Banks Islands (Totemism in Polynesia and Melanesia, in J.A.I., XXXIX, p. 172). Among the Arunta it is the spirit of an ancestor who is believed to impregnate the mother; in the Banks Islands, it is the spirit of an animal or vegetable, as the theory supposes. But as the ancestral spirits of the Arunta have an animal or vegetable form, the difference is slight. Therefore we have not mentioned it in our exposition.

[602] Social Origins, London, 1903, especially ch. viii, entitled The Origin of Totem Names and Beliefs, and The Secret of the Totem, London, 1905.

[603] In his Social Origins especially, Lang attempts to reconstitute by means of conjecture the form which these primitive groups should have; but it seems superfluous to reproduce these hypotheses, which do not affect his theory of totemism.

[604] On this point, Lang approaches the theory of Julius Pickler (see Pickler and Szomolo, Der Ursprung des Totemismus. Ein Beitrag zur materialistirchen Geschichtstheorie, Berlin, 36 pp. in 8vo). The difference between the two hypotheses is that Pickler attributes a higher importance to the pictorial representation of the name than to the name itself.

[605] Social Origins, p. 166.

[606] The Secret of the Totem, p. 121; cf. pp. 116, 117.

[607] The Secret of the Totem, p. 136.

[608] J.A.I., Aug., 1888, pp. 53-54; cf. Nat. Tr., pp. 89, 488, 498.

[609] "With reverence," as Lang says (The Secret of the Totem, p. 111).

[610] Lang adds that these taboos are the basis of exogamic practices.

[611] Ibid., p. 125.

[612] However, we have not spoken of the theory of Spencer. But this is because it is only a part of his general theory of the transformation of the ancestor-cult into the nature-cult. As we have described that already, it is not necessary to repeat it.

[613] Except that Lang ascribes another source to the idea of the great gods: as we have already said, he believes that this is due to a sort of primitive revelation. But Lang does not make use of this idea in his explanation of totemism.

[614] For example, in a Kwakiutl myth, an ancestral hero pierces the head of an enemy by pointing a finger at him (Boas, Vth Rep. on the North. Tribes of Canada, B.A.A.S., 1889, p. 30).

[615] References supporting this assertion will be found on p. 128, n. 1, and p. 320, n. 1.

[616] See Bk. III, ch. ii.

[617] See, for example, Howitt, Nat. Tr., p. 482; Schürmann, The Aboriginal Tribes of Port Lincoln, in Woods, Nat. Tr. of S. Australia, p. 231.

[618] Frazer has even taken many facts from Samoa which he presents as really totemic (See Totemism, pp. 6, 12-15, 24, etc.). It is true that we have charged Frazer with not being critical enough in the choice of his examples, but so many examples would obviously have been impossible if there had not really been important survivals of totemism in Samoa.

[619] See Turner, Samoa, p. 21 and ch. iv and v.

[620] Alice Fletcher, A Study of the Omaha Tribe, in Smithsonian Rep. for 1897, pp. 582 f.

[621] Dorsey, Siouan Sociology, in XVth Rep., p. 238.

[622] Ibid., p. 221.

[623] Riggs and Dorsey, Dakota-English Dictionary, in Contrib. N. Amer. Ethnol., VII, p. 508. Many observers cited by Dorsey identify the word wakan with the words wakanda and wakanta, which are derived from it, but which really have a more precise signification.

[624] XIth Rep., p. 372, § 21. Miss Fletcher, while recognizing no less clearly the impersonal character of the wakanda, adds nevertheless that a certain anthropomorphism has attached to this conception. But this anthropomorphism concerns the various manifestations of the wakanda. Men address the trees or rocks where they think they perceive the wakanda, as if they were personal beings. But the wakanda itself is not personified (Smithsonian Rep. for 1897, p. 579).

[625] Riggs, Tah-Koo Wah-Kon, pp. 56-57, quoted from Dorsey, XIth Rep., p. 433, § 95.

[626] XIth Rep., p. 380, § 33.

[627] Ibid., p. 381, § 35.

[628] Ibid., p. 376, § 28; p. 378, § 30; cf. p. 449, § 138.

[629] Ibid., p. 432, § 95.

[630] Ibid., p. 431, § 92.

[631] Ibid., p. 433, § 95.

[632] Orenda and a Definition of Religion, in American Anthropologist, 1902, p. 33.

[633] Ibid., p. 36.

[634] Tesa, Studi del Thavenet, p. 17.

[635] Boas, Kwakiutl, p. 695.

[636] Swanton, Social Condition, etc, of the Tlinkit Indians, XXVIth Rep., 1905, p. 451, n. 2.

[637] Swanton, Contributions to the Ethnology of the Haida, p. 14; cf. Social Condition, etc., p. 479.

[638] In certain Melanesian societies (Banks Islands, North New Hebrides) the two exogamic phratries are found which characterize the Australian organization (Codrington, The Melanesians, pp. 23 ff.). In Florida, there are regular totems, called butos (ibid., p. 31). An interesting discussion of this point will be found in Lang, Social Origins, pp. 176 ff. On the same subject, and in the same sense, see W. H. R. Rivers, Totemism in Polynesia and Melanesia, in J.A.I., XXXIX, pp. 156 ff.

[639] The Melanesians, p. 118, n. 1. Cf. Parkinson, Dreissig Jahre in der Südsee, pp. 178, 392, 394, etc.

[640] An analysis of this idea will be found in Hubert and Mauss, Théorie Générale de la Magie, in Année Sociol., VII, p. 108.

[641] There are not only totems of clans but also of guilds (A. Fletcher, Smithsonian Rep. for 1897, pp. 581 ff.).

[642] Fletcher, op. cit., pp. 578 f.

[643] Ibid., p. 583. Among the Dakota, the totem is called Wakan. See Riggs and Dorsey, Dakota Grammar, Texts and Ethnol., in Contributions N. Amer. Ethn., 1893, p. 219.

[644] James's Account of Long's Expedition in the Rocky Mountains, I, p. 268. (Quoted by Dorsey, XIth Rep., p. 431, § 92.)

[645] We do not mean to say that in principle every representation of religious forces in an animal form is an index of former totemism. But when we are dealing with societies where totemism is still apparent, as is the case with the Dakota, it is quite natural to think that these conceptions are not foreign to it.

[646] See below, same book, ch. ix, § 4, pp. 285 ff.

[647] The first spelling is that of Spencer and Gillen; the second, that of Strehlow.

[648] Nat. Tr., p. 548, n. 1. It is true that Spencer and Gillen add: "The idea can be best expressed by saying that an Arungquiltha object is possessed of an evil spirit." But this free translation of Spencer and Gillen is their own unjustified interpretation. The idea of the arungquiltha in no way implies the existence of spiritual beings, as is shown by the context and Strehlow's definition.

[649] Die Aranda, II, p. 76, n.

[650] Under the name Boyl-ya (see Grey, Journal of Two Expeditions, II, pp. 337-338).

[651] See above, p. 42. Spencer and Gillen recognize this implicitly when they say that the arungquiltha is a "supernatural force." Cf. Hubert and Mauss, Théorie Générale de la Magie, in Année Sociol., VII, p. 119.

[652] Codrington, The Melanesians, pp. 191 ff.

[653] Hewitt, loc. cit., p. 38.

[654] There is even ground for asking whether an analogous notion is completely lacking in Australia. The word churinga, or tjurunga as Strehlow writes, has a very great similarity, with the Arunta. Spencer and Gillen say that it designates "all that is secret or sacred. It is applied both to the object and to the quality it possesses" (Nat. Tr., p. 648, s.v. churinga). This is almost a definition of mana. Sometimes Spencer and Gillen even use this word to designate religious power or force in a general way. While describing a ceremony among the Kaitish, they say that the officiant is "full of churinga," that is to say, they continue, of the "magic power emanating from the objects called churinga." Yet it does not seem that the notion of churinga has the same clarity and precision as that of the mana in Melanesia or of the wakan among the Sioux.

[655] Yet we shall see below (this book, ch. viii and ix) that totemism is not foreign to all ideas of a mythical personality. But we shall show that these conceptions are the product of secondary formations: far from being the basis of the beliefs we have just analysed, they are derived from them.

[656] Loc. cit., p. 38.

[657] Rep. Peabody Museum, III, p. 276, n. (quoted by Dorsey, XIth Rep., p. 435).

[658] See above, p. 35.

[659] In the expressions such as Ζεὺς ὕει or Ceres succiditur, it is shown that this conception survived in Greece as well as in Rome. In his Götternamen, Usener has clearly shown that the primitive gods of Greece and Rome were impersonal forces thought of only in terms of their attributes.

[660] Définition du phénomène religieux, in Année Sociol., II, pp. 14-16.

[661] Preanimistic Religion, in Folk-Lore, 1900, pp. 162-182.

[662] Ibid., p. 179. In a more recent work, The Conception of Mana (in Transactions of the Third International Congress for the History of Religions, II, pp. 54 ff.), Marrett tends to subordinate still further the animistic conception of mana, but his thought on this point remains hesitating and very reserved.

[663] Ibid., p. 168.

[664] This return of preanimism to naturism is still more marked in Clodd, Preanimistic Stages of Religion (Trans. Third Inter. Congress for the H. of Rel., I, p. 33).

[665] Théorie générale de la Magie, in Année Sociol., VII, pp. 108 ff.

[666] Der Ursprung der Religion und Kunst, in Globus, 1904, Vol. LXXXVI, pp. 321, 355, 376, 389; 1905, Vol. LXXXVII, pp. 333, 347, 380, 394, 413.

[667] Globus, LXXXVII, p. 381.

[668] He clearly opposes them to all influences of a profane nature (Globus, LXXXVI, p. 379a).

[669] It is found even in the recent theories of Frazer. For if this scholar denies to totemism all religious character, in order to make it a sort of magic, it is just because the forces which the totemic cult puts into play are impersonal like those employed by the magician. So Frazer recognizes the fundamental fact which we have just established. But he draws different conclusions because he recognizes religion only where there are mythical personalities.

[670] However, we do not take this word in the same sense as Preuss and Marrett. According to them, there was a time in religious evolution when men knew neither souls nor spirits: a preanimistic phase. But this hypothesis is very questionable: we shall discuss this point below (Bk. II, ch. viii and ix).

[671] On this same question, see an article of Alessandro Bruno, Sui fenomeni magico-religiosi della communità primitive, in Rivista italiana di Sociologia, XII Year, Fasc. IV-V, pp. 568 ff., and an unpublished communication made by W. Bogoras to the XIV Congress of the Americanists, held at Stuttgart in 1904. This communication is analysed by Preuss in the Globus, LXXXVI, p. 201.

[672] "All things," says Miss Fletcher, "are filled with a common principle of life," Smiths. Rep. for 1897, p. 579.

[673] Hewitt, in American Anthropologist, 1902, p. 36.

[674] The Melanesians, pp. 118-120.

[675] Ibid., p. 119.

[676] See above, p. 103.

[677] Pickler, in the little work above mentioned, had already expressed, in a slightly dialectical manner, the sentiment that this is what the totem essentially is.

[678] See our Division du travail social, 3rd ed., pp. 64 ff.

[679] Ibid., p. 76.

[680] This is the case at least with all moral authority recognized as such by the group as a whole.

[681] We hope that this analysis and those which follow will put an end to an inexact interpretation of our thought, from which more than one misunderstanding has resulted. Since we have made constraint the outward sign by which social facts can be the most easily recognized and distinguished from the facts of individual psychology, it has been assumed that according to our opinion, physical constraint is the essential thing for social life. As a matter of fact, we have never considered it more than the material and apparent expression of an interior and profound fact which is wholly ideal: this is moral authority. The problem of sociology—if we can speak of a sociological problem—consists in seeking, among the different forms of external constraint, the different sorts of moral authority corresponding to them and in discovering the causes which have determined these latter. The particular question which we are treating in this present work has as its principal object, the discovery of the form under which that particular variety of moral authority which is inherent in all that is religious has been born, and out of what elements it is made. It will be seen presently that even if we do make social pressure one of the distinctive characteristics of sociological phenomena, we do not mean to say that it is the only one. We shall show another aspect of the collective life, nearly opposite to the preceding one, but none the less real (see p. 212).

[682] Of course this does not mean to say that the collective consciousness does not have distinctive characteristics of its own (on this point, see Représentations individuelles et représentations collectives, in Revue de Métaphysique et de Morale, 1898, pp. 273 ff.).

[683] This is proved by the length and passionate character of the debates where a legal form was given to the resolutions made in a moment of collective enthusiasm. In the clergy as in the nobility, more than one person called this celebrated night the dupe's night, or, with Rivarol, the St. Bartholomew of the estates (see Stoll, Suggestion und Hypnotismus in der Völkerpsychologie, 2nd ed., p. 618, n. 2).

[684] See Stoll, op. cit., pp. 353 ff.

[685] Ibid., pp. 619, 635.

[686] Ibid., pp. 622 ff.

[687] The emotions of fear and sorrow are able to develop similarly and to become intensified under these same conditions. As we shall see, they correspond to quite another aspect of the religious life (Bk. III, ch. v).

[688] This is the other aspect of society which, while being imperative, appears at the same time to be good and gracious. It dominates us and assists us. If we have defined the social fact by the first of these characteristics rather than the second, it is because it is more readily observable, for it is translated into outward and visible signs; but we have never thought of denying the second (see our Règles de la Méthode Sociologique, preface to the second edition, p. xx, n. 1).

[689] Codrington, The Melanesians, pp. 50, 103, 120. It is also generally thought that in the Polynesian languages, the word mana primitively had the sense of authority (see Tregear, Maori Comparative Dictionary, s.v.).

[690] See Albert Mathiez, Les origines des cultes révolutionnaires (1789-1792).

[691] Ibid., p. 24.

[692] Ibid., pp. 29, 32.

[693] Ibid., p. 30.

[694] Ibid., p. 46.

[695] See Mathiez, La Théophilanthropie et la Culte décadaire, p. 36.

[696] See Spencer and Gillen, Nor. Tr., p. 33.

[697] There are even ceremonies, for example, those which take place in connection with the initiation, to which members of foreign tribes are invited. A whole system of messages and messengers is organized for these convocations, without which the great solemnities could not take place (see Howitt, Notes on Australian Message-Sticks and Messengers, in J.A.I., 1889; Nat. Tr., pp. 83, 678-691; Spencer and Gillen, Nat. Tr., p. 159; Nor. Tr., p. 551).

[698] The corrobbori is distinguished from the real religious ceremonies by the fact that it is open to women and uninitiated persons. But if these two sorts of collective manifestations are to be distinguished, they are, none the less, closely related. We shall have occasion elsewhere to come back to this relationship and to explain it.

[699] Except, of course, in the case of the great bush-beating hunts.

[700] "The peaceful monotony of this part of his life," say Spencer and Gillen (Nor. Tr., p. 33).

[701] Howitt, Nat. Tr., p. 683. He is speaking of the demonstrations which take place when an ambassador sent to a group of foreigners returns to camp with news of a favourable result. Cf. Brough Smyth, I, p. 138; Schulze, loc. cit., p. 222.

[702] See Spencer and Gillen, Nat. Tr., pp. 96 f.; Nor. Tr., p. 137; Brough Smyth, II, p. 319.—This ritual promiscuity is found especially in the initiation ceremonies (Spencer and Gillen, Nat. Tr., pp. 267, 381; Howitt, Nat. Tr., p. 657), and in the totemic ceremonies (Nor. Tr., pp. 214, 298, 237). In these latter, the ordinary exogamic rules are violated. Sometimes among the Arunta, unions between father and daughter, mother and son, and brothers and sisters (that is in every case, relationship by blood) remain forbidden (Nat. Tr., pp. 96 f.).

[703] Howitt, Nat. Tr., pp. 535, 545. This is extremely common.

[704] These women were Kingilli themselves, so these unions violated the exogamic rules.

[705] Nor. Tr., p. 237.

[706] Nor. Tr., p. 391. Other examples of this collective effervescence during the religious ceremonies will be found in Nat. Tr., pp. 244-246, 365-366, 374, 509-510 (this latter in connection with a funeral rite). Cf. Nor. Tr., pp. 213, 351.

[707] Thus we see that this fraternity is the logical consequence of totemism, rather than its basis. Men have not imagined their duties towards the animals of the totemic species because they regarded them as kindred, but have imagined the kinship to explain the nature of the beliefs and rites of which they were the object. The animal was considered a relative of the man because it was a sacred being like the man, but it was not treated as a sacred being because it was regarded as a relative.

[708] See below, Bk. III, ch i, § 3.

[709] At the bottom of this conception there is a well-founded and persistent sentiment. Modern science also tends more and more to admit that the duality of man and nature does not exclude their unity, and that physical and moral forces, though distinct, are closely related. We undoubtedly have a different conception of this unity and relationship than the primitive, but beneath these different symbols, the truth affirmed by the two is the same.

[710] We say that this derivation is sometimes indirect on account of the industrial methods which, in a large number of cases, seem to be derived from religion through the intermediacy of magic (see Hubert and Mauss, Théorie générale de la Magie, Année Sociol., VII, pp. 144 ff.); for, as we believe, magic forces are only a special form of religious forces. We shall have occasion to return to this point several times.

[711] At least after he is once adult and fully initiated, for the initiation rites, introducing the young man to the social life, are a severe discipline in themselves.

[712] Upon this particular aspect of primitive societies, see our Division du travail social, 3rd ed., pp. 123, 149, 173 ff.

[713] We provisionally limit ourselves to this general indication: we shall return to this idea and give more explicit proof, when we speak of the rites (Bk. III).

[714] On this point, see Achelis, Die Ekstase, Berlin, 1902, especially ch. i.

[715] Cf. Mauss, Essai sur les variations saisonnières des sociétés eskimos, in Année Sociol., IX, p. 127.

[716] Thus we see how erroneous those theories are which, like the geographical materialism of Ratzel (see especially his Politische Geographie), seek to derive all social life from its material foundation (either economic or territorial). They commit an error precisely similar to the one committed by Maudsley in individual psychology. Just as this latter reduced all the psychical life of the individual to a mere epiphenomenon of his physiological basis, they seek to reduce the whole psychical life of the group to its physical basis. But they forget that ideas are realities and forces, and that collective representations are forces even more powerful and active than individual representations. On this point, see our Représentations individuelles et représentations collectives, in the Revue de Métaphysique et de Morale, May, 1898.

[717] See above, pp. 188 and 194.

[718] Even the excreta have a religious character. See Preuss, Der Ursprung der Religion und Kunst, especially ch. ii, entitled Der Zauber der Defäkation (Globus, LXXXVI, pp. 325 ff.).

[719] This principle has passed from religion into magic: it is the totem ex parte of the alchemists.

[720] On this point see Règles de la méthode sociologique, pp. 5 ff.

[721] Procopius of Gaza, Commentarii in Isaiam, 496.

[722] See Thévenot, Voyage au Levant, Paris, 1689, p. 638. The fact was still round in 1862.

[723] Lacassagne, Les Tatouages, p. 10.

[724] Lombroso, L'homme criminel, I, p. 292.

[725] Lombroso, ibid., I, pp. 268, 285, 291 f.; Lacassagne, op. cit., p. 97.

[726] See above, p. 127.

[727] For the authority of the chiefs, see Spencer and Gillen, Nat. Tr., p. 10; Nor. Tr., p. 25; Howitt, Nat. Tr., pp. 295 ff.

[728] At least in Australia. In America, the population is more generally sedentary; but the American clan represents a relatively advanced form of organization.

[729] To make sure of this, it is sufficient to look at the chart arranged by Thomas, Kinship and Marriage in Australia, p. 40. To appreciate this chart properly, it should be remembered that the author has extended, for a reason unknown to us, the system of totemic filiation in the paternal line clear to the western coast of Australia, though we have almost no information about the tribes of this region, which is, moreover, largely a desert.

[730] The stars are often regarded, even by the Australians, as the land of souls and mythical personages, as will be established in the next chapter: that means that they pass as being a very different world from that of the living.

[731] Op. cit., I, p. 4. Cf. Schulze, loc. cit., p. 243.

[732] Of course it is to be understood that, as we have already pointed out (see above, p. 155), this choice was not made without a more or less formal agreement between the groups that each should take a different emblem from its neighbours.

[733] The mental state studied in this paragraph is identical to the one called by Lévy-Bruhl the law of participation (Les fonctions mentales dans les sociétés inférieures, pp. 76 ff.). The following pages were written when this work appeared and we publish them without change; we confine ourselves to adding certain explanations showing in what we differ from M. Lévy-Bruhl in our understanding of the facts.

[734] See above, p. 230.

[735] Another cause has contributed much to this fusion; this is the extreme contagiousness of religious forces. They seize upon every object within their reach, whatever it may be. Thus a single religious force may animate the most diverse things which, by that very fact, become closely connected and classified within a single group. We shall return again to this contagiousness, when we shall show that it comes from the social origins of the idea of sacredness (Bk. III, ch. i, in fine).

[736] Lévy-Bruhl, op. cit., pp. 77 ff.

[737] Ibid., p. 79.

[738] See above, p. 146.

[739] This is the case with the Gnanji; see Nor. Tr., pp. 170, 546; cf. a similar case in Brough Smyth, II, p. 269.

[740] Australian Aborigines, p. 51.

[741] There certainly was a time when the Gnanji women had souls, for a large number of women's souls still exist to-day. However, they never reincarnate themselves; since in this tribe the soul animating a new-born child is an old reincarnated soul, it follows from the fact that women's souls do not reincarnate themselves, that women cannot have a soul. Moreover, it is possible to explain whence this absence of reincarnation comes. Filiation among the Gnanji, after having been uterine, is now in the paternal line: a mother no longer transmits her totem to her child. So the woman no longer has any descendants to perpetuate her; she is the finis familiæ suæ. To explain this situation, there are only two possible hypotheses; either women have no souls, or else they are destroyed after death. The Gnanji have adopted the former of these two explanations; certain peoples of Queensland have preferred the latter (see Roth, Superstition, Magic and Medicine, in N. Queensland Ethnog., No. 5, § 68).

[742] "The children below four or five years of age have neither soul nor future life," says Dawson. But the fact he thus relates is merely the absence of funeral rites for young children. We shall see the real meaning of this below.

[743] Dawson, p. 51; Parker, The Euahlayi, p. 35; Eylmann, p. 188.

[744] Nor. Tr., p. 542; Schürmann, The Aboriginal Tribes of Port Lincoln, in Woods, p. 235.

[745] This is the expression used by Dawson, p. 50.

[746] Strehlow, I, p. 15, n. 1; Schulze, loc. cit., p. 246; this is the theme of the myth of the vampire.

[747] Strehlow, I, p. 15; Schulze, p. 244; Dawson, p. 51. It is true that it is sometimes said that souls have nothing corporeal; according to certain testimony collected by Eylmann (p. 188), they are ohne Fleisch und Blut. But these radical negations leave us sceptical. The fact that offerings are not made to the souls of the dead in no way implies, as Roth thinks (Superstition, Magic, etc., § 65), that they do not eat.

[748] Roth, ibid., § 65; Nor. Tr., p. 530. It sometimes happens that the soul emits odours (Roth, ibid., § 68).

[749] Roth, ibid., § 67; Dawson, p. 51.

[750] Roth, ibid., § 65.

[751] Schürmann, Aborig. Tr. of Port Lincoln, in Woods, p. 235.

[752] Parker, The Euahlayi, pp. 29, 35; Roth, ibid., §§ 65, 67, 68.

[753] Roth, ibid., § 65; Strehlow, I, p. 15.

[754] Strehlow, I, p. 14, n. 1.

[755] Frazer, On Certain Burial Customs, as Illustrative of the Primitive Theory of the Soul, in J.A.I., XV, p. 66.

[756] This is the case with the Kaitish and the Unmatjera; see Nor. Tr., p. 506; and Nat. Tr., p. 512.

[757] Roth, ibid., §§ 65, 66, 67, 68.

[758] Roth, ibid., § 68; this says that when someone faints after a loss of blood, it is because the soul is gone. Cf. Parker, The Euahlayi, p. 38.

[759] Parker, The Euahlayi, pp. 29, 35; Roth, ibid., § 65.

[760] Strehlow, I, pp. 12, 14. In these passages he speaks of evil spirits which kill little children and eat their souls, livers and fat, or else their souls, livers and kidneys. The fact that the soul is thus put on the same plane as the different viscera and tissues and is made a food like them shows the close connection it has with them. Cf. Schulze, p. 245.

[761] For example, among the peoples on the Pennefather River (Roth, ibid., § 68), there is a name for the soul residing in the heart (Ngai), another for the one in the placenta (Cho-i), and a third for the one which is confounded with the breath (Wanji). Among the Euahlayi, there are three or even four souls (Parker, The Euahlayi, p. 35).

[762] See the description of the Urpmilchima rite among the Arunta (Nat. Tr., pp. 503 ff.).

[763] Spencer and Gillen, Nat. Tr., pp. 497 and 508.

[764] Nor. Tr., pp. 547, 548.

[765] Ibid., pp. 506, 527 ff.

[766] Meyer, The Encounter Bay Tribe, in Woods, p. 198.

[767] Nor. Tr., pp. 551, 463; Nat. Tr., p. 553.

[768] Nor. Tr., p. 540.

[769] Among the Arunta and Loritja, for example (Strehlow, I, p. 15, n. 2; II, p. 77). During life, the soul is called gumna, and ltana after death. The ltana of Strehlow is identical with the ulthana of Spencer and Gillen (Nat. Tr., pp. 514 ff.). The same is true of the tribes on the Bloomfield River (Roth, Superstition, etc., §66).

[770] Eylmann, p. 188.

[771] Nat. Tr., pp. 524, 491, 496.

[772] Nor. Tr., pp. 542, 504.

[773] Mathews, Ethnol. Notes on the Aboriginal Tribes of N.S. Wales and Victoria, in Journal and Proc. of the Roy. Soc. of N.S. Wales, XXXVIII, p. 287.

[774] Strehlow, I, pp. 15 ff. Thus, according to Strehlow, the dead live in an island in the Arunta theory, but according to Spencer and Gillen, in a subterranean place. It is probable that the two myths coexist and are not the only ones. We shall see that even a third has been found. On this conception of an island of the dead, see Howitt, Nat. Tr., p. 498; Schürmann, Aborig. Tr. of Port Lincoln, in Woods, p. 235; Eylmann, p. 189.

[775] Schulze, p. 244.

[776] Dawson, p. 51.

[777] In these same tribes evident traces of a more ancient myth will be found, according to which the dead live in a subterranean place (Dawson, ibid.).

[778] Taplin, The Narrinyeri, pp. 18 f.; Howitt, Nat. Tr., p. 473; Strehlow, I, p. 16.

[779] Howitt, Nat. Tr., p. 498.

[780] Strehlow, I, p. 16; Eylmann, p. 189; Howitt, Nat. Tr., p. 473.

[781] These are the spirits of the ancestors of a special clan, the clan of a certain poisonous gland (Giftdrüsenmänner).

[782] Sometimes the work of the missionaries is evident. Dawson speaks of a real hell opposed to paradise; but he too tends to regard this as a European importation.

[783] Dorsey, XIth Rep., pp. 419-420, 422, 485. Cf. Marillier, La survivance de l'âme et l'idée de justice chez les peuples non-civilisés, Rapport de l'Ecole des Hautes Études, 1893.

[784] They may be doubled temporarily, as we shall see in the next chapter: but these duplications add nothing to the number of the souls capable of reincarnation.

[785] Strehlow, I, p. 2.

[786] Nat. Tr., p. 73, n. 1

[787] On this set of conceptions, see Nat. Tr., pp. 119, 123-127, 387 ff.; Nor. Tr., pp. 145-174. Among the Gnanji, it is not necessarily near the oknanikilla that the conception takes place. But they believe that each couple is accompanied in its wanderings over the continent by a swarm of souls of the husband's totem. When the time comes, one of these souls enters the body of the wife and fertilizes it, wherever she may be (Nor. Tr., p. 169).